bq: Why is it possible then to alter replicationFactor via MODIFYCOLLECTION in the collections API
Because MODIFYCOLLECTION just changes properties in the collection definition generically and replicationFactor just happens to be one. IOW there's no overarching reason. It would be extra work to dis-allow that one case and possibly introduce errors without changing any functionality so nobody was willing to put in the effort. Best, Erick On Thu, Apr 13, 2017 at 5:48 AM, Shawn Heisey <apa...@elyograg.org> wrote: > On 4/13/2017 3:22 AM, Johannes Knaus wrote: >> Ok. Thank you for your quick reply. Though I still feel a little >> uneasy. Why is it possible then to alter replicationFactor via >> MODIFYCOLLECTION in the collections API? What would be the use case >> for this parameter at all then? > > If you use a very specific storage method for your indexes -- HDFS -- > then replicationFactor has meaning beyond initial collection creation, > in conjunction with the "autoAddReplicas" feature. > > https://cwiki.apache.org/confluence/display/solr/Running+Solr+on+HDFS#RunningSolronHDFS-AutomaticallyAddReplicasinSolrCloud > > If you are NOT utilizing the very specific HDFS storage engine, then > everything you were told applies. With standard storage mechanisms, > replicationFactor has zero meaning after initial collection creation, > and changing the value will have no effect. > > Thanks, > Shawn >
